<?xml version="1.0" standalone="yes"?>
<Paper uid="H05-2011">
  <Title>DialogueView: an Annotation Tool for Dialogue</Title>
  <Section position="3" start_page="0" end_page="20" type="intro">
    <SectionTitle>
3 UtteranceView
</SectionTitle>
    <Paragraph position="0"> The annotations in WordView are utilized in building the next view, UtteranceView. This view shows the utterances of two speakers as if it were a script for a movie.</Paragraph>
    <Paragraph position="1"> To derive a single ordering of the utterances of the two  speakers, we use the start time of each utterance as annotated in WordView. We refer to this process as linearizing the dialogue (Heeman and Allen, 1995). The order of the utterances should show how the speakers are sequentially adding to the dialogue, and is our motivation for defining utterances as being small enough so that they are not affected by subsequent speech of the other speaker.</Paragraph>
    <Paragraph position="2"> Users can annotate utterance tags in UtteranceView besides WordView. WordView is more suitable for tags that depend on the exact timing of the words, or a very local context, such as whether an utterance is abandoned or incomplete, or whether there is overlap speech. UtteranceView is more suitable for tags that relate the utterance to other utterances in the dialogue, such as whether an utterance is an answer, a statement, a question, or an acknowledgment. Whether an annotation tag can be used in WordView or UtteranceView (or both) is specified in the configuration file. Which view a tag is used in does not affect how it is stored in the annotation files (also in XML format).</Paragraph>
    <Paragraph position="3"> In UtteranceView, users can annotate hierarchical groupings of utterances. We call each grouping a block, and blocks can have other blocks embedded inside of them. Each block is associated with a summary, which users need to fill in. Blocks can be closed; when a block is closed, it is replaced by its summary, which is displayed as if it were said by the speaker who initiated the block.</Paragraph>
    <Paragraph position="4"> Just as utterances can be tagged, so can discourse blocks.</Paragraph>
    <Paragraph position="5"> The block tags scheme is also specified in the configuration file.</Paragraph>
    <Paragraph position="6"> UtteranceView supports two types of playback. The first playback simply plays both channels mixed, which is exactly what is recorded. The second playback is slightly different. It takes the linearization into account and dynamically builds an audio file in which each utterance in turn is concatenated together, and a 0.5 second pause is inserted between each utterance. This gives the user an idealized rendition of the utterances, with overlapping speech separated. By comparing these two types of playbacks, users can aurally check if their linearization of the dialogue is correct.</Paragraph>
    <Paragraph position="7"> Users can use the configuration file to customize UtteranceView. Typically, UtteranceView gives users a clean display of what is going on in a dialogue. This clean display removes reparanda and editing terms in speech repairs, and it also removes abandoned speech, which has no contributions to the conversation.1 UtteranceView also supports adding texts or symbols to an utterance based on the tags, such as adding &amp;quot;?&amp;quot; after a question, &amp;quot;...&amp;quot; after an incomplete utterance, and &amp;quot;+&amp;quot; at both the beginning and end of an overlapping utterance to signal the overlap. (c.f. Childes scheme (MacWhinney, 2000)).</Paragraph>
    <Paragraph position="8"> 1Note that these clean processes are optional. Users can specify them in the configuration file.</Paragraph>
  </Section>
class="xml-element"></Paper>